Supercharge Your Innovation With Domain-Expert AI Agents!

Sight line estimation model establishment method and device, electronic equipment and storage medium

A line-of-sight estimation and model-building technology, applied in the field of artificial intelligence, can solve problems such as low accuracy of the line-of-sight estimation model, difficulty in obtaining user facial images, etc., and achieve the effect of improving accuracy

Active Publication Date: 2022-07-08
BEIHANG UNIV
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, in practical applications, it is difficult to obtain a large number of user facial images containing gaze direction annotations, resulting in low accuracy of the trained gaze estimation model

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Sight line estimation model establishment method and device, electronic equipment and storage medium
  • Sight line estimation model establishment method and device, electronic equipment and storage medium
  • Sight line estimation model establishment method and device, electronic equipment and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0035] figure 2 A schematic flowchart of a method for establishing a line of sight estimation model provided in Embodiment 1 of the present application, the method includes the following steps:

[0036] S101. Iteratively execute the following processing until the current first rotation distribution loss satisfies the first constraint condition: rotate the first user face image at multiple first angles to obtain multiple first rotated images, and the first user face image includes Line-of-sight direction labeling; using the plurality of first rotated pictures as the input of the line-of-sight estimation model, according to the plurality of first line-of-sight estimation results, the first angle and the line-of-sight direction label output by the line-of-sight estimation model, through the first line of sight estimation model output a calculation of the loss of rotation distribution, and performing constraint adjustment on the line-of-sight estimation model;

[0037] S102. Per...

Embodiment 2

[0069] Image 6 A schematic structural diagram of an apparatus for establishing a line-of-sight estimation model provided in Embodiment 2 of the present application, such as Image 6 As shown, the device includes:

[0070] The pre-training module 61 is configured to iteratively perform the following processing until the current first rotation distribution loss satisfies the first constraint condition: rotating the first user face picture at multiple first angles to obtain multiple first rotated pictures, the A user's face picture includes line-of-sight direction annotations; the plurality of first rotated pictures are used as the input of the line-of-sight estimation model, according to the plurality of first line-of-sight estimation results, the first angle and the line of sight output from the line-of-sight estimation model Orientation annotation, through the calculation of the first rotation distribution loss, the constraint adjustment is performed on the line of sight est...

Embodiment 3

[0103] Figure 7 is an apparatus block diagram of an apparatus for establishing a line of sight estimation model according to an exemplary embodiment, the apparatus may be a mobile phone, a computer, a digital broadcasting terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, Personal digital assistants, etc.

[0104] Device 800 may include one or more of the following components: processing component 802, memory 804, power supply component 806, multimedia component 808, audio component 810, input / output (I / O) interface 812, sensor component 814, and communication component 816.

[0105] The processing component 802 generally controls the overall operation of the device 800, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 802 can include one or more processors 820 to execute instructions to perform all or some of the steps of the met...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention provides a sight line estimation model establishment method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out the iteration of a sight line estimation model for multiple times based on the calculation of rotation distribution loss according to a picture obtained through rotating a user face picture containing a sight line direction label at multiple angles, and carrying out the constraint adjustment, so as to complete the pre-training; after the pre-training is completed, according to a picture obtained by rotating a user face picture which does not contain a sight line direction label at multiple angles, on the basis of calculation of a pseudo sight line direction label and rotation distribution loss provided by a pseudo label generation model, iterating a pseudo label generation model and a sight line estimation model for multiple times to carry out constraint adjustment; and completing the training of the sight line estimation model. According to the scheme, a large number of training samples can be obtained to train the sight line estimation model based on a small number of pictures containing sight line direction labels, so that the accuracy of the sight line estimation model is improved.

Description

technical field [0001] The present application relates to the technical field of artificial intelligence, and in particular, to a method, apparatus, electronic device and storage medium for establishing a line of sight estimation model. Background technique [0002] Line of sight is an important clue to reveal the way people understand the external environment. Gaze estimation techniques have been widely used in fields such as human-computer interaction, virtual reality, augmented reality, and medical analysis. Line-of-sight estimation techniques have attracted a lot of attention in recent years. The line of sight estimation technology refers to a technology that calculates the direction of the user's line of sight through the captured image of the user's face. [0003] Currently, state-of-the-art techniques typically use convolutional neural networks to solve the line-of-sight estimation problem. The input is a user's face image captured by a color camera, and the output...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06F30/20G06F3/01G06F111/04
CPCG06F30/20G06F3/013G06F2111/04
Inventor 陆峰鲍屹伟
Owner BEIHANG UNIV
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com